INSIDE EOSC 02 – Sy Holsinger
Andrew Dubber
Hi, and welcome to Inside EOSC, a podcast all about the inner workings of the European Open Science Cloud. I’m Andrew Dubber. I’m a professor in innovation, a senior researcher at the Industry Commons Foundation, and a consortium member of an EOSC project called LUMEN—and for the purpose of this episode, also GRAPHIA.
Each month, I introduce—with no apologies, maybe with a terrible pun—Luminaries: people who are central to the LUMEN project and to the wider EOSC ecosystem. And today’s guest, CTO of the OPERAS, who we’re going to talk about a little more in just a moment, and also the project coordinator of GRAPHIA—Sy Holsinger. Sy, thanks so much for joining us.
Sy Holsinger
Yeah, thanks for having me.
Andrew Dubber
What is OPERAS?
Sy Holsinger
So OPERAS is a European research infrastructure that’s focused on open scholarly communication in social sciences and humanities.
Andrew Dubber
And a CTO in that context does what?
Sy Holsinger
Well, pretty much I’m in charge of all of the services, the digital services that they offer as part of the research infrastructure.
Andrew Dubber
You’ve been involved in EU projects for a little while now.
Sy Holsinger
Yeah, since about 2007. So I think that’s roughly 18 years or so now, I guess.
Andrew Dubber
Is it a French accent I’m hearing? German?
Sy Holsinger
Yeah. So originally from the US and then been living in Europe for almost 20 years now.
Andrew Dubber
Right. Okay. Interesting piece of your background.US military. You were Air Force.
Sy Holsinger
Yeah. And that was the whole reason I got sent to Italy, in which I remained for a large portion of that time.
Andrew Dubber
Stayed for love? Is that the story or?
Sy Holsinger
Well, the majority of these kinds of stories normally start there.
Andrew Dubber
Right. Fantastic. Okay. So the European projects you’ve been involved in—obviously there’s been some intersection with EOSC along the way.
You’ve had some experience with EOSC. I’ve asked this to a few people, and I get different answers every time I ask the question: what’s EOSC?
Sy Holsinger
So the European Open Science Cloud—I mean, there’s a decent history there. So I was kind of there before EOSC even started. So I mean, I started in distributed computing, starting with grid computing, which was really developed to support the large amounts of data that were being produced by the Large Hadron Collider.
So a number of different projects that were being coordinated by CERN. And then really, like many outputs from CERN, they start with an individual research objective and the output ends up becoming useful for a number of different areas. And in this case, the distributed computing infrastructure that they built ended up being useful for other scientific disciplines, other than the high-energy physics, which was the main focus.
So it started moving into biology and chemistry and life sciences and a number of other multidisciplinary disciplines. So they wanted to set up basically some type of legal entity or coordination that could sustain these 250-some data centres that were being federated across more than 30 countries. And then that’s where the EGI Foundation was established in 2010, which then I joined as part of doing some of their policy, business development, service management—things like that.
Andrew Dubber
I’m pausing there for just a second because the European context is a world of acronyms. EGI—what are we talking about?
Sy Holsinger
Well, like many things, it started as an acronym, and right now is no longer an acronym because they branded it basically as just EGI. But it did start as the European Grid Infrastructure, which has its roots in what was being developed through those projects coordinated by CERN. But the grid then evolves, and many people will probably know more—cloud computing is a little bit more popular across the mainstream. So they cloud infrastructures. So it basically just became a distributed infrastructure rather than an acronym anymore.
Andrew Dubber
Right. And before OPERAS, you were working for EGI.
Sy Holsinger
And then, so that’s where my kind of story continued, let’s say, in the e-infrastructure world. And I ended up spending roughly 11 years at EGI in a number of different roles. But EGI doesn’t live alone in the e-infrastructure world.
There are other e-infrastructures, for example, EUDAT, which does data management; GÉANT, which offers, started with networking and coordinating all the national research networks, OpenAIRE as well as another major player in the e-infrastructure world. So the story here is basically this is kind of where the EOSC was born—that you had a number of different e-infrastructures, and you had individual research infrastructures that were more focused in specific scientific disciplines. So then you’d research infrastructures that were on top of them.
So in life sciences, there was ELIXIR, not to name them all. So really, the initial idea, at least according to my opinion—many people might have other opinions on what the EOSC is and isn’t. But for me, the original idea was: let’s take a researcher perspective.
So you have all of these research infrastructures, which are providing services, ultimately with the goal to provide researchers with useful things in order for them to conduct research—and even more advanced research than otherwise wouldn’t be done without these e-infrastructures. So if you take the researcher perspective, you want to make it easy for them to access services. But when you have a multitude of different e-infrastructures, you then ultimately start sending researchers into all kinds of different places in order to find the different services that you need.
So the idea of the EOSC is how can we kind of bring these together in a more harmonised environment to try to make it easier for researchers to access these services. So one of the original EOSC projects, which was called EOSC Hub, and that was coordinated by EGI, was kind of this first initial step. So it was EGI together with EUDAT to try to bring together some of these services in a more harmonised fashion.
So that went through a couple of years’ phase, and then EOSC Hub turned into EOSC Future, and then that added additional e-infrastructures into the mix to continuously try to bring these together. And now the EOSC is just a series of continuation of this initiative—how can we bring research infrastructures together with the e-infrastructures and try to make access to these digital services easier for the researchers? But obviously, this is complex, because federations already in themselves are complex.
So if you had a federation together with a federation, you have federations of federations, you know. So and that itself is a grand endeavour, and I almost have to tip my cap to even trying to take this on.
Andrew Dubber
Well, so I guess that explains why I get a different answer every time I ask what EOSC is, because it’s really complicated. There’s a lot going on—a lot of moving parts. But the word infrastructure comes up a lot.
And when I hear infrastructure, I think of roads and buildings and public services, and so on. But it’s about computing, and it’s about storage, and it’s about processing, and these sorts of things. But is there a thing that you can point to and say “that’s EOSC”, or is it just a kind of a nebulous consortium of people kind of broadly working in the same direction?
Is there a rulebook, basically?
Sy Holsinger
I mean, actually, they just tried to—I think they just released the EOSC handbook to try to explain what it is. But I mean, fundamentally, it’s like anything that it starts as an initial idea. It has an acronym—European Open Science Cloud.
But then cloud—what does that, you know. Does it have its history in cloud computing, et cetera? But it’s really EOSC has just kind of turned into almost an idea, right? It’s basically: how can we bring a very fragmented, complex research environment into a system? And even in the early days of EOSC, they called it the “system of systems”.
Andrew Dubber
Yeah.
Sy Holsinger
Right. And just trying to make sense of everything that has been funded—you know, fundamentally, this has been publicly funded for years, and you want to just make sure that the things that have been funded are being used.
So how can you make it useful, which is already a complex environment, in a little bit more of a simplistic way?
Andrew Dubber
The things that OPERAS seem to be interested in—like open access publication, cultural heritage, social sciences. I mean, social sciences has the word “science” in it, but it still seems a little peripheral to what I imagine from what you’ve described EOSC is about, which is the hardcore computing, the massive calculations, the spread of infrastructure. Where’s the intersection of OPERAS and EOSC?
Sy Holsinger
So I think it’s, you know, everything kind of goes in phases, and things get built up in time. So, of course, EOSC has its origins in a lot of the STEM related sciences—let’s say the harder sciences—but it never is going to be limited to it. So it’s like ever-expanding, with additional disciplines and things.
So I think OPERAS kind of brings this new dimension, new discipline, which comes with new challenges. And this is the social sciences and humanities publishing, you know, but more open access publishing. But even the term “publishing” was being used in, you know, early drafts of the EOSC strategy, the SRIA document, because you’re talking about publishing in the terms of publishing scientific publications, but not publishing as in the publishing world, which is much broader than that.
So I think OPERAS, together with some of its other partners in social sciences and humanities, brings the topic of scholarly communication to add to the existing scientific disciplines. So for me, this is just a natural evolution of EOSC is to make sure that it’s covering the breadth of scientific disciplines that exist in the world. So it’s not necessarily it was just limited by, it’s just you can’t cover everything all at once.
So you have to start it somewhere. So now, scholarly communication has been included in one of the EOSC terminology—it’s “opportunity area”. So this is kind of a new opportunity area where the EOSC is organising around the topic of scholarly communication, so that we can feed into the next versions of the strategy documents.
What is EOSC? How does that feel? And so we’re being onboarded, let’s say, into it.
Andrew Dubber
When you Google Sy Holsinger, you see that you are a member of the EOSC expert area in Open Scholarly Communication. You’ve just described it as an “opportunity area”, unless that’s a different thing. I was going to ask you: what is an expert area? But what is this area, and how are they organised?
Sy Holsinger
So really, there’s a number of different EOSC-related support projects that go in trying to support the whatever community they’re trying to involve, and then how to ensure that gets captured with inside of whatever structures that EOSC put into place.
So I was just one of the, let’s say, early representatives at the beginning to even say, “Hey, we’re here, we should be included, this is how it should work.” And then you get coordinated together with representatives from the EOSC, then representatives from the Scholarly Communication and Scholarly Publications.
So right now, one of my colleagues, Sona Arasteh, is really kind of leading a lot of that activity now.
Andrew Dubber
OPERAS is not so much an organisation, but a group of organisations. Do I understand that correctly?
Sy Holsinger
Yeah, it’s a federated infrastructure, distributed infrastructure.
Andrew Dubber
Is it a member organisation, presumably?
Sy Holsinger
A member organisation, for right now. It’s an individual institute to join OPERAS as a member. Of course, we’re coordinated by a legal entity that’s based in Brussels, but that’s just to provide the underlying coordination and management of it.
But it’s really a distributed organisation, where the members themselves are participating in the activities. And really, they’re the ones that bring in the main expert knowledge, and then carry out a lot of the activities that need to be done across the infrastructure.
Andrew Dubber
And that explains why OPERAS is a member of, say, the LUMEN project and the GRAPHIA project, but also, so are members of OPERAS.
Sy Holsinger
Yeah, so that’s the thing is you end up having at least one organisation that can kind of start, or coordinate, or lead. But we never do things by ourselves because, you know, OPERAS, we’re a small organisation that’s coordinating a much larger community. And so, normally, whenever we try to participate in things, it’s bringing that community along with us.
Andrew Dubber
So you’re not a named member of the LUMEN project, but you are the CTO of an organisation that is a member of the LUMEN project. And there’s an overlap, like all the conversations I have about LUMEN currently seem to be very interested in knowledge graphs and particularly the history with what was the GoTriple platform and how that’s being extended. Now, the project that you are named—I mean, you’re the project coordinator for—is the GRAPHIA project.
So again, knowledge graphs, GoTriple. There seems to be a kind of very strong Venn diagram overlap between these two projects. You want to explain the difference?
Sy Holsinger
Yeah, so basically, as a role of CTO, it’s basically looking at the service portfolio as a whole. And this kind of goes into two areas. So, the first one is looking at our current service portfolio and how do we want to develop, evolve, improve, and that over time.
And a lot of that is through participation in a number of research projects. So, the overlap here with GoTriple, which is one of our flagship services, is being funded by a number of different initiatives. So, I’m always interested in understanding what we are developing, what is improving.
And of course, this gets distributed across a number of different projects, whether or not I’m formally involved or not. But the interest is that LUMEN is developing, improving GoTriple as its service, as well as GoTriple itself is part of the GRAPHIA project, where then looking at the other side of it, which is GRAPHIA, and where this idea came from. And this was, instead of looking at the current service portfolio and developing our current services, this was kind of our discussion about looking forward.
What’s next? Where would we like to go? And we kind of felt like knowledge graphs were an opportunity for us to add a new service, potentially, to the service portfolio. But fundamentally, it was something that could be useful for the broader SSH community.
In fact, that’s why we call it the SSH Knowledge Graph, and it will be integrating a number of existing Knowledge Graphs that already exist. So, that was kind of a future opportunity that we saw with GRAPHIA
Andrew Dubber
Right. It sounds like you’re very involved in connecting things that are already quite complicated.
Sy Holsinger
Yeah, it’s pretty much.
Andrew Dubber
I mean, I was going to ask you—I mean, I understand how you go from the Air Force to living in Europe. I’m not sure how you go from the Air Force to living in the world of social sciences and humanities, and open publication.
Sy Holsinger
Like many mother-tongue English speakers living abroad, you start as an English teacher. But going back to university days, I started in communications and then added business management to the field of studies. So, when opportunities came up, it was more of a communications and marketing division with inside of an IT company, and they were doing project management of EC-funded projects and were looking for project managers and things. So that was kind of my first introduction, but was more with my communications hat on.
So I started in dissemination and communication, event organisation and stuff. But with the series of projects that were being funded by CERN, there was also the component of having outreach to industry. So, trying to bridge research with industry to increase exploitation of research outputs, et cetera.
So that’s what kind of led to the evolution of one role. And you get into policy, business development, then I get into service management and how to manage distributed infrastructures. And that’s kind of how one thing kind of led to the next.
But it’s this kind of putting pieces of a puzzle, a very complex puzzle together and trying to make sense of it. So, and then I realised that trying to make sense of a complex environment and where you could organise it, structure it, and then be able to explain it in a way that people can understand ended up being a skill in and of itself, basically. So that would lead to the next.
That’s why I really enjoyed the opportunity at OPERAS, because building out the infrastructure at EGI over 11 some years, OPERAS offers that opportunity to kind of establish very similar things, but even if it’s in a very scientific discipline. But we’ve learned that the entire community, European community, has learned a lot over this kind of federated, multidisciplinary, multi-organisational structure. So there’s a lot of repeatable mechanics that you can apply to the system, but it’s always going to be complex, and there’s always going to be a lot of missing parts. And I think I’ve just tend to come to enjoy it, whether it’s a masochist or not, it’s fun at some point.
Andrew Dubber
The thing about knowledge graphs, which I find really interesting, and you’re deeply immersed in the world of knowledge graphs and ontologies. Now, I was talking to Yann Le Franc on another episode of this podcast, and he said, “I came to this from neuroscience. So, of course, I’m interested in knowledge graphs”.
I talked to Julien Homo, who’s the technical coordinator. He said, “I came to this from computer science and AI. So, of course, I’m interested in knowledge graphs”.
What’s your, of course, “I’m interested in knowledge graphs”?
Sy Holsinger
So, I’m a pragmatic guy, and when you’re working a lot, and you spend such a large percentage of your life in work, I want to do something useful. I’m in the utility business. If I don’t see that we’re doing something that has value, I’m not interested.
So, I really want to produce something that actually has value. So, if I look at this from a problem-solving perspective, basically, especially in social science and humanities, that data is fragmented—where it’s being stored, and how are researchers able to access this data? And then you realise that there’s a wealth of knowledge that’s being produced by all of these researchers from across the world, and they’re all stored in different repositories and different databases, and things like that.
So, then when you think about a knowledge graph, you’re like, okay, so I can tie in all of these different repositories, which essentially is knowledge, right? And we can pull that together, and then we can also connect or relate that knowledge to each other, and then give researchers even a more useful tool for them to conduct research, generate new knowledge in a much more easier, complex way. So I think knowledge graphs itself is a way that we could try to simplify access to fragmented data and knowledge.
So I think that’s a pretty cool, worthwhile use of time, let’s say.
Andrew Dubber
Like me, you’re from somewhere else. And so, as an outsider to a certain extent—even though you’ve been here a long time, as have I—can you identify something that is quintessentially European about these sorts of frameworks and infrastructures, and approaches to knowledge? And the European Open Science Cloud doesn’t have an equivalent, as far as I know, elsewhere in the same way—in terms of its federated nature, in terms of its ambitions and goals and how it works.
What’s European about the European Open Science Cloud?
Sy Holsinger
I think it’s an interesting question. For the last several years, there have been these large-scale research projects that have happened in the US, so they are happening, right? But I think the biggest difference is basically the collaboration that then continues from the research that’s being funded—one’s leading into the next.
I also think, in the let’s say, early 2010s, the European Commission really started pushing this concept of sustainability. And now, we take it for granted that, basically, of course, you talk about sustainability. But that wasn’t always the case.
Andrew Dubber
By sustainability, you mean that it can continue, not that we’re recycling our plastics?
Sy Holsinger
Yeah, exactly. Yeah, that’s the other important clarification, even in terms of how we think about the term. But it was, and it’s correct.
It’s basically saying: we are funding something through a project. You will achieve some result. We don’t want that result to die, because if not, it’s a wasted use of taxpayer money.
So, it’s a very important question that they started to ask, which is: no, this result needs to continue. It needs to be sustained. Yeah.
And I think that was a kind of a challenge for Europe, but it was something not really being addressed from what I could see in the US. But even in Europe, that became quite a challenge, because fundamentally, you were turning a bunch of researchers and research institutes into service providers. And that’s actually what got me into service management, into looking at standards and frameworks for how this could be organised, because that’s not the training that the researchers and the research institutes had.
But they were fundamentally being forced into being service providers, because they had to sustain these services. So, that whole community or market that got created, you know, I think that’s fundamentally unique to Europe—which is we are establishing legal entities in order to sustain research outputs. And I think this is one of the biggest differences in Europe than the US, because they aren’t really pushing legal entities or sustained organisations that then can keep this going.
And then there are some collaborations between different universities. Those are kind of almost self-organised between US universities to try to keep this stuff going. But it’s not systemic the way that it is here.
And then that really reinforces these kind of communities, but collaborations between a lot of these institutes, because many of them get created, but then they have to work together. And that, in itself, continues this kind of cycle of a constant series of collaborations between them. And that’s where I think you can probably see how something like the EOSC could even be created, because there was this underlying system of organisations trying to sustain research outputs.
So they took a European approach to try to keep these collaborations between them going. And that to me is fundamentally European from my point of view.
Andrew Dubber
I think, from the perspective of GoTriple, I think that’s a really fantastic thing, because something that was as good as that—because GoTriple was a really phenomenal outcome of a project—to take that further and have projects that build on existing work and carry this idea further. Absolutely fantastic. Dark side of this might be that there becomes a culture of throwing good money after bad.
If there wasn’t a good outcome of the GoTriple project, I mean, it turned into something that actually wasn’t any good, but then it throws millions more euros into sort of digging the hole deeper. Is there a lot of that? Do you see that?
Sy Holsinger
No, I don’t think any specific examples come to my mind. I can appreciate that if you invest into something, it becomes seen as your baby, your child, and you want to see it to go away. But I think things come down to—I don’t want to say the word indicators or KPIs, the key performance indicators—but what it really means is: is it being used?
And I think that’s the question. So, if you’re having an honest conversation about whether something’s being used, the community or the researchers who you’re trying to serve, they’ll be the ultimate ones that keep things alive. Because if it’s not being used, and they’re using something else, then certain things will just die over time.
Obviously, you don’t want to see certain outputs not continuing into the future. But for me, I think that’s okay. If you’re truly innovating and you’re truly trying to solve the problems of whoever your target customers are—in this case, let’s say it’s the researchers—then they’ll use it because you’re solving a problem that you want.
So if it’s being used and then you continue to innovate in terms of what they need, then the service will stay. But I admit that that’s a risk that many organisations and even many different sectors can fall into. And I think that’s a testament to the LUMEN project, to the GRAPHIA project, is to evolve the service that you have into the way that the researchers will continue it.
We had another project called FASCA, which is basically trying to introduce GoTriple into the researcher’s workflow. So, they’re already using a number of different, let’s say, data visualisation tools and other different tools, but they really are into workflows.
So, if you can insert your product into the researcher workflow, then again, it just increases its utility and increases its utilisation.
And that’s how I think you can avoid just throwing good money after bad.
Andrew Dubber
Well, speaking of increase of utility, there is a presupposition in a lot of this stuff— particularly around EOSC, but let’s say for the LUMEN project—that giving researchers and scientists a whole lot of tools that help them do their work better is a good thing. So, that what? To what end?
Are we talking about progress for humanity? Are we talking about curing cancer? Are we talking about, you know, what is the KPI for better research?
Sy Holsinger
This is an interesting discussion, because it kind of brings out another area where OPERAS is trying to participate in, and this is in the area of research assessment. And having, you know, this general conversation that’s happening across the research world is, how are we measuring or assessing research? You know, because, you know, traditionally it was just going into these high-impact journals, you know, and all of these numbers and questioning that.
And so, for us, for OPERAS, the interest in research assessment is because the diversity of research outputs that come from the SSH community—you know, open access books and, you know, monographs, and the list goes on. And it’s not always this traditional—what you would find in the STEM cells is scientific publications going into high-impact journals. So I think there’s a very interesting conversation.
Obviously, OPERAS is just part of that conversation. We’re involved in a project called GraspOS that is looking at it. You know, there’s the CoARA Working Group that’s looking at that.
So, yeah, we do need to, as an entire research community, reassess how we are assessing the research that’s coming from, you know, this push for open access publication and diversity of research outputs.
Andrew Dubber
Is utility the leading value in there, or are there other values that are being pushed forward in terms of research assessment?
Sy Holsinger
I think it’s the generation of new knowledge that, I think, ultimately kind of pushes society and humanity as a whole. You know, asking questions, you know, questioning things, or raising new questions, and, you know, just kind of advancing human society as a whole. I think it is good for everybody.
Andrew Dubber
So, to what extent is computing implicated in this progress of humanity?
Sy Holsinger
Well, computing is, I mean, even if you just look at GRAPHIA and the knowledge graph, but really, fundamentally, you know, artificial intelligence is, you know, AI is becoming a little bit everywhere, and SSH community is not immune to the introduction of AI. So we’re really kind of trying to coordinate and take the pulse of the community in terms of how they feel about the use of AI, the introduction of AI into SSH activities. So, historically, SSH isn’t producing the amount of data that you might be able to compare to the Large Hadron Collider and the high-energy physics—petabytes and beyond worth of data.
But AI is computationally heavy, you know—the usage of GPUs. And so, there is a computational need, even if you’re not talking about large amounts, large amounts of data. So, really, any type of AI processing is going to require compute.
So, what we’re seeing that now, even in the SSH community, as we get into things like knowledge graphs, chatbots, anything trying to produce natural language processing requires higher levels of compute. So, now we find ourselves, you know, partnering with infrastructure providers like our Polish provider, PSNC, to try to come up with computing infrastructure to satisfy our needs. So, even SSH is entering into the larger demand for computing infrastructure, even if it’s not for large amounts of storage.
But we’ll see, as, you know, the knowledge graph continues to introduce other scientific-discipline-specific knowledge graphs, that’s going to increase the computing—or even just the storage—that we need to be able to manage the data that we’re talking about. And the more that it gets used, the more that they can integrate it, the bigger it gets. And the infrastructure requirements are always going to increase.
So, but not necessarily a bad thing. It’s just one of the many components that go into trying to provide these kind of advanced digital services for researchers.
Andrew Dubber
Right. I heard a great line recently, somebody said, “If humanities aren’t in the conversation, then all you get is the inhumanities.” But I feel like, I mean, every conversation today is a conversation about AI in one way or another—or at least it seems like that.
But every AI conversation is a conversation ultimately about ethics. And we always end up there. We always end up in conversations about ownership and bias and, you know, all these other things.
And it feels like humanities have got a big thing to say about that. And maybe it’s not just AI for humanities, but also humanities for AI. And to what extent is that part of the LUMEN/GRAPHIA agenda?
Sy Holsinger
I mean, taking a step back, just even thinking about this, there could be philosophical questions in here. And what’s more SSH than philosophical questions about what does this mean for humanity? And even just exploring the impact of this is an SSH, you know, endeavour.
So, we’re already embedded as soon as you mention it, because there’s technical as well as non-technical implications of this. What effect will this have on society? How will this change the new generation, you know, with how younger kids are using AI even within their homework?
And that’s going to be the new generation. So, there are a lot of societal impacts that we can look at just on its usage in addition to, you know, the technical implications of this. But I can at least say with inside of GRAPHIA, we’re taking this very seriously.
So, we have a dedicated activity that’s dedicated to ethics and legal aspects. So, as questions arise from, for example, even just in the technical architecture, when you’re designing the knowledge graph, it’s already natural questions have already been raised. So what we try to do is use that activity to capture all of these kind of legal and ethical questions that are arising even from before doing any development.
Just even looking at the high-level architecture of what we may or may not do is automatically eliciting these kinds of ethical questions. So we, at least, are trying to collect them, ask the questions, collect them, and then try to find a way to at least answer them. And then you can go even further and say, OK, answer them for your own project.
But these are questions that many others are going to be asking them themselves. So, inside of OPERAS, we have different working groups on different topics. And we’ve recently launched, we call them special interest groups.
So, if there’s a topic that the community feels particularly strongly about, we capture that topic, and then we pull the people from the community who are interested in it. And they run meetings, have discussions, produce even white papers that then can be made public as an output from the exercise. So, we recently launched a special interest group specifically on AI.
So, we can have this conversation, and as well as kind of make—not necessarily just make more public, but make more and more people aware—that these are the questions. This is how we’re currently addressing them. Could even get feedback on that.
And that, to me, I think feeds out into the larger conversation. So we just want to contribute to that conversation because everybody’s asking themselves the same questions.
Andrew Dubber
So, I want to cycle back to something you said early on in the conversation about AI, where you grouped together legal and ethical considerations. I find that really interesting, because I work with a lawyer who says that he’s not interested in ethics. It’s not what he does.
He’s a lawyer, and there’s a joke in there somewhere, but also the idea that don’t look for ethics in the law. The law says what the law says, and a lawyer’s job is to ensure compliance, maybe. But the ethical consideration needs to be done outside of the legal framework, and then can be brought to be the basis of the legal framework.
Is there a kind of , I guess, a problem in bundling those things together? So, we think of ethics as just being compliance with whatever the rules happen to be.
Sy Holsinger
I don’t know. I don’t see, I see them working kind of hand in hand. Like you almost can’t have one without the other.
Of course, you’re going to have different expertise in those different fields. Maybe the challenge becomes: it is bridging that gap between the two. But that’s why I think most of the language—I mean, outside even the GRAPHIA project, even in conversations with insider OPERAS—and we’re looking at what are the topics that we want to address in the future.
Whenever it comes to AI, we bundle them together. Ethics with… So, I’m not saying it’s an easy solve, because, as you’re right, lawyers are going to be lawyers. But I don’t think you can have AIs without ethics.
But then, you know, there are different ways that you can enforce ethics. Maybe this is another philosophical question that we can have. But still, you know, legal stuff can help us try to maybe enforce some of the things that we, as a society, collectively agree we don’t want to see happen.
And, you know, unfortunately, this is kind of sometimes how society works. It gets civilised, gets governed by the rules and laws that we put into place. So, if it happens while we’re driving a car, I don’t see any difference in applying that to how we want to introduce artificial intelligence into our lives.
Andrew Dubber
Right. Well, I’m not going to ask you to solve AI and the future of humanity in an AI world. But I am interested in what you think a good outcome of these sorts of projects within the framework of EOSC or related to EOSC, you know, if you’re working in social sciences and humanities, and you’re intersecting with a project like LUMEN or GRAPHIA, and you’re putting knowledge graphs into these knowledge domains—how will we know we’ve done a good job?
Sy Holsinger
Everything’s multifaceted. Every research project that we ever get in, there are non-technical and technical elements to this. So, the non-technical side will be contributing to the overall society dialogue in terms of how we are, you know, operating ourselves as a society.
And then there’ll be the technical implementations, which come back to our previous reason for doing what we do, and that’s in delivering value to the researchers that want to use them. So, for me, you know, the end result of these projects is that we have identified the results that we think need to be delivered. And ultimately, if we deliver those, those will be used.
And if they’re being used, we could consider the projects a success.
Andrew Dubber
Fantastic. Sy Holsinger, thanks so much for your time. Really appreciate you joining us on the podcast.
Sy Holsinger
Thanks for having me.
Andrew Dubber
That’s Sy Holsinger, CTO of OPERAS, who are partners on the EOSC LUMEN project. And he’s also the coordinator of the GRAPHIA project. And that’s Inside EOSC, a podcast from LUMEN, which receives funding from the European Union’s Horizon Research and Innovation Programme.
I’m Andrew Dubber. I’m back next month with an interview with another EOSC Insider. You can subscribe to Inside EOSC wherever you listen to podcasts.
Thanks for listening.